The Evolution of Party Organizations in Europe:
The Three Faces of Party Organization
Richard S. Katz, The Johns Hopkins University
Peter Mair, University of Leiden
At least since the beginning of the 1980s, much of the writing on the
strategies, tactics, and policies of parties has explicitly cautioned against the
treatment of party as a “unitary actor” (e.g., Daalder 1983; Laver and Schofield 1990). Party leaders, it is now argued, may differ from one another and
from party followers with regard to the ends which they pursue and the
resources which they employ; even within the leadership itself, it is suggested, the conflict between factions may be such as to militate against any
theories taking the party as a whole as the relevant unit of analysis. Nonetheless, there is a striking lack of consensus regarding the number and type
of different units into which a party may be disaggregated, and the extent to
which this disaggregation may be applied to an understanding of processes
of organizational change.
When parties are disaggregated in the organizational literature, writers
still tend to fall back on the simple division between leaders and followers
on which Michels based his influential law of oligarchy. This dividing line
is, of course, sometimes qualified or refined, and more nuanced distinctions
are occasionally drawn between “the party in government” and “the party in
the electorate,” or between “the parliamentary party” and the “extraparliamentary party,” and so on. Attention is also sometimes drawn to
separate groups of “activists” or “middle-level elites,” that is, to groups
mediating, or simply located, between “the leaders” and “the followers.”
Despite these nuances, however, when party is broken down in organizational terms, the process now, as before, seems to hinge on a single hierarchy (e.g. Duverger 1951; Kirchheimer 1966).
Even when one moves beyond simple dichotomies, the unidimensionality of these distinctions remains a serious problem. For example, the simple
leader-follower dichotomy fails to distinguish between those party leaders
who are in public office (in parliament or in government) and those party
______________
RICHARD S. KATZ is Professor of Political Science at The Johns Hopkins University.
PETER MAIR is Professor of Political Science at the University of Leiden.
The American Review of Politics, Vol. 14, Winter, 1993: 593-617
©1993 The American Review of Politics
594 | Richard S. Katz and Peter Mair
leaders who instead are based in the party’s own “private” offices. Both are
party “leaders” within the same party; their interests, however, and hence
their interaction with “the followers,” need not coincide. In a similar and
overlapping sense, the parliamentary versus extra-parliamentary dichotomy
not only ignores the fact that the parliamentary organization may be divided
into those members who are actually in government and those who are not
(e.g., Andeweg 1992), but also fails to recognize that the extra-parliamentary
party may be divided between its own leadership, on the one hand, and the
ordinary members (as well as the activists and middle-level elites, and so
on), on the other. Panebianco does allude to this theme, both in his reference
to the potential conflict between a party’s national apparatus and its
peripheral organizations (1988, 58) and in his discussion of the tension
between “internal leaders” and the parliamentary group (1988, 173), but in
neither case is this developed into a more comprehensive attempt to analyze
organizational change and adaptation.
Three Faces of Party Organization
It is our contention that rather than analyzing parties according to a
simple parliamentary versus extra-parliamentary dichotomy, or a simple
leaders versus followers hierarchy (no matter how finely subdivided), it is
more productive to consider parties as being comprised of a number of different elements, or faces, each of which potentially interacts with all of the
others. Although each face may itself be quite heterogeneous, and a full
analysis of party organization would require that these faces be disaggregated and analyzed, we believe it to be possible to identify subsystems within parties that, by virtue of their location in the party and the wider political
system, will interact with one another in understandably patterned ways.
As a first approximation, we propose consideration of three faces of
party organization. The first is the party in public office, e.g., in parliament
or government. The second is the party on the ground, that is the members,
activists, and so on. The third is the party central office, that is, the national
leadership of the party organization which, at least in theory, is organizationally distinct from the party in public office, and which, at the same time,
organizes and is usually representative of the party on the ground.
In an earlier stage of developing these ideas, we suggested a different
set of three faces to characterize party organizations: the party as governing
organization; the party as membership organization; and the party as bureaucratic organization. Our argument was that each face entails a different set of
resources, constraints, opportunities, and patterns of motivation that bear on
party leaders based within it, and that because of these differences, there
Evolution of Party Organizations in Europe | 595
would be conflicts among party leaders (beyond those simply engendered by
personal ambition) on the basis of which organizational change could be
understood.
In this revision, we retain the basic argument, but now believe we see a
more effective way of dividing the same pie. The “party in public office”
takes the place of what we originally called the “party as governing organization.” Although this is primarily a change in nomenclature, it underlines
the fact that even parties that are in opposition usually have leaders who
occupy public office in parliament, in regional and local councils, and so
forth.
The “party on the ground” and the “party central office” were, in the
old trichotomy, combined into the “party as membership organization.” On
the one hand, as noted above, the party central office has sometimes been
treated as if it were annexed to, and hence also absorbed within, the party in
public office, thus allowing the simple distinction between a unitary set of
party leaders and a unitary set of party followers. Such a perspective thereby
tends to attribute to the party central office the incentives and resources
which really apply primarily to the party in public office. In addition, this
perspective tends to end up by treating the extra-parliamentary party as if it
were without leadership or organization, and as if the party congress, for
example, were no more than a mob or a mass-meeting.
On the other hand, and more frequently, the party central office has
been equated with the party on the ground, thus permitting a distinction
between a homogeneous parliamentary party and a homogeneous extraparliamentary party. Not only does this force us to ignore important aspects
of each of these two faces, however, but as well it forces us to ignore
tensions between the party on the ground and the party central office. Yet, it
is precisely these tensions that give rise to accusations of oligarchic
tendencies within political parties. Indeed, Michels’ whole theory fails
unless we can distinguish these two faces.
Finally, rather than aggregating the party bureaucracy into an independent element of party organization, we now put more emphasis on the
fact that parties often have several separate bureaucracies. Thus, in the
present conceptualization, we disaggregate the party bureaucracies into parts
associated with each of the three faces.
The Party in Public Office
The key feature of the party in public office is that, at least in democratic countries, it is dominated by those who have themselves been successful in elections, and who depend on continued electoral success in order to
596 | Richard S. Katz and Peter Mair
keep their positions. If one of the defining characteristics of parties is they
“bas[e] claims of legitimacy on electoral success” (Katz 1987, 8), then the
party in public office is the quintessential core, and the outward symbol of
success, of party writ large.
The rewards or goals pursued by members of the party in public office
may be of several types. First in much of the theorizing about parties (e.g.,
Downs 1957), if not necessarily in the minds of party actors, are the personal
rewards of office. Aside from material benefits, these include the psychic
rewards of power and status. As pointed out in the literature, these rewards
are divisible and transferable only to a limited extent, and fundamentally
accrue only to the particular individuals holding office.
Downsian assumptions that suggest policy is only a means to the end
of office notwithstanding, a second set of goals is undoubtedly (in most
cases) the pursuit of particular policy objectives. Here the distinction between the party in public office and other faces of party is not that these
rewards are uniquely available to the party in public office, but rather that
the actors in this face of the party are uniquely positioned to have a personal
role in their achievement. While one consequence might be to give these
goals greater immediacy, another is likely to be a greater appreciation of the
constraints and limitations on policy making. In this respect, members of the
party in public office are more likely to see compromise as incremental
movement toward a desired goal rather than as partial retreat from a correct
position.
An important characteristic of the party in public office is its transience, with continued corporate existence and individual membership
dependent on extra-party (i.e., electoral) forces. Moreover, although some of
the rewards of office, as well as the capacity to influence the course of
public policy, are available to members of the opposition, the total stock and
value of individual rewards is far greater for the party(s) in government. And
this, as well, is ultimately in the hands of the voters, rather than the party
itself.
The need to win elections, both in order to remain in office and to pursue effectively the other rewards that attracted them to politics in the first
place is the first important constraint on members of the party in public
office. This means that they must be attentive not only to the electorate, but
as well to those who control the resources necessary for a successful election
campaign.
A second constraint is the obligations of government. Although one
might argue that some parties have been devoid of any sense of civic responsibility, and it is unreasonable to assume that self-interested politicians
undergo some kind of apotheosis on achieving office, government responsi-
Evolution of Party Organizations in Europe | 597
bility, and perhaps to a lesser extent the public responsibilities of being the
“loyal opposition,” do appear to constrain most parties in government. This
undoubtedly is in part for moral reasons, but there are also reasons of a more
practical nature. Parties in government must expect to be held electorally accountable for the general condition of the country (Lewis-Beck
1988), and thus have an incentive to care for the general welfare. Governing
brings members of the party in public office into regular contact, and mutual
dependence, with members of the higher civil service whose views about the
nature of the national interest and the proper role of government are likely to
be far less partisan and short-term than their own (Aberbach et al. 1981);
both the need to develop a cooperative relationship with members of the
civil service and the simple social pressure inherent in working with them
are likely to foster a sense of general responsibility on the part of the politicians. Moreover, even if responsibility does not constrain officials, the fact
that they must work cooperatively—with coalition partners, civil servants,
officials at other levels of government—if they are to be effective limits the
freedom of action of members of the party in public office.
At the same time, being in office gives the party in public office a
number of important resources that can be used in internal party politics. The
most obvious is that the members of the party in public office have the legal
authority to make governmental decisions; they vote on the bills, direct the
bureaucrats, and so forth. One class of such decisions relates to a second
resource available to the party in public office: patronage. A third set of
resources is time, expertise and information. The members of the party in
public office often are paid salaries that allow them to devote full time to
politics; their positions give them experience and expertise; and, moreover,
they have access to the expertise and information gathering and processing
capabilities of the state bureaucracy. Finally, the party in public office has
the legitimacy conferred by a public mandate.
The Party on the Ground
In the case of parties with formal mass memberships, the members are
the basis of the party on the ground, but more loosely it can be taken to
include the core of regular activists, financial supporters, and even loyal
voters, whether or not they are formally enrolled as party “members.” The
key characteristics of this face of party are voluntary membership, permanence, and regularity. Although there may be various requirements for joining and maintaining formal membership, entry and exit are, for the most
part, based on the private choices of the individual members. For most parties, both the scale and the intermittent participation of the average member
598 | Richard S. Katz and Peter Mair
require representative institutions. While the primary locus of the party on
the ground is, of course, diffused throughout the country, it is manifested
organizationally at the national level by the party congress, and at various
other levels by other committees and congresses, with established rules to
fix the number and types of officials, their competence and terms, etc.
Although there may be some individual incentives for membership and
activity in the party on the ground—for example, the local party office may
serve various social functions for its members, local leadership positions
may confer some status, activity may put the member in line for rewards of
patronage or nomination to office (and thus, if successful, membership in the
party in public office)—the primary incentives for members of the party on
the ground are public purposive (policy), symbolic, and solidaristic. Thus,
making and adhering to formal statements of party policy and identity are
likely to be of great significance, and this may put the party on the ground in
conflict with the party in public office. Moreover, while members of the
party on the ground will certainly see winning elections as preferable to
losing, the sacrifices they are prepared to make for that end may be quite
limited.
The party on the ground has a variety of resources. Most typically, they
have their own labor, which can be important both for election campaigns
and other political propaganda/agitation and also for filling the variety of
positions on local governing and advisory boards that are allocated to or won
by the party. The party on the ground can be a source of other electorally
important resources, especially money and votes. They also bring local
knowledge to the party, in some cases augmented by a paid staff. In some
conceptions of democracy, the party on the ground also has its own special
legitimacy as the political embodiment of the segment of society that the
party as a whole claims to represent.
The most important constraint on the party on the ground is simply that
they are not the party in public office, and consequently are unable to make
governmental decisions themselves. A second constraint applies not so much
to the party on the ground as a whole as to its leadership, and that is that the
party on the ground is generally a voluntary organization, from which exit is
always a viable option. Leaders of this face must, therefore, satisfy their
members not only to retain their positions of leadership, but also to maintain
an organization to lead.
The Party Central Office
Generally located in the national capital, the party central office consists of two (frequently overlapping) groups of people, the national executive
Evolution of Party Organizations in Europe | 599
committee or committees, and the central party staff or secretariat. Members
of the first group may be recruited in a variety of ways. Some may be
elected by the party congress, or in some other way appear to represent the
party on the ground; others may be representatives or leaders of the party in
public office; still others may be representatives of ancillary or affiliated
organizations. In many cases, not only will the top party bureaucrat be an exofficio member of the national executive, but (s)he may appoint several
other officials who become ex-officio members as well. In other words,
despite appearances, the national executive of a party may be less a representative body than a self-perpetuating and autonomous element of the overall party structure.
In principle, the central party bureaucracy should be the servant of the
national executive, but it may have many resources (not least of which may
be ex-officio members on the national executive) to support a more assertive
role than “servant” implies. Thus, one might ask for any party whether the
national executive or the party bureaucracy is the dominant force in the
central office. In some cases, indeed, the party bureaucracy may be the true
central office, with the national executive reduced to a purely nominal or
ceremonial role.
The primary resources of the party central office are its centrality,
expertise, and formal position at the apex of the party organization. To these
might be added that many of its members are leaders of other faces of the
party. This is an ambiguous situation, however. On one hand, it means that
when it is united, the party central office can draw on the resources of the
other faces to establish itself as the dominant locus of decision within the
party. But, on the other hand, when its members are not united, their status
as leaders of the other faces can transform the party central office from the
dominant locus of decision into merely a battle-ground, or alternatively an
empty shell that is both impotent and largely ignored.
It is, of course, true for each of the faces of party that its strength is
maximized only when its individual members are united. This is a particularly important constraint on the party central office, however, because the
individual motivations of its members and the individual level constraints on
them are likely to be more disparate. In particular, the members of the party
executive are likely to owe their positions to different faces of party, and
have to maintain the support of their individual constituencies if they are to
remain in the central office. For example, members representing the party on
the ground should value ideological purity more, while those representing
the party in public office value electoral victory more. Security is likely to
be extremely important to party bureaucrats, who even more than the party
in public office are virtually by definition individuals who live from rather
600 | Richard S. Katz and Peter Mair
than for politics, but it is the security of their own positions within the party
hierarchy rather than the security of the party’s position within the political
system that should be their most immediate concern.
Another constraint on the party central office is that it, even more than
the party on the ground, which may be taken to include (or be fused with)
the holders of local political office, cannot make and implement public
policy on its own. Indeed, in a sense, the key question regarding the party
central office is not to detail its resources and the constraints it faces, but
rather to ask why a party would have a central office at all. From this perspective, the key resource of the central office is its unique ability to perform
the functions assigned to it, and the constraint is that to the degree to which
those functions lose their value or alternative means of performing them are
found, that resource is devalued.
If the defining characteristic of a party is that it is attempting to win
power, there is no need to ask why a party would value its public office
holding face; it is the raison d’être of the whole enterprise. Various studies
(Katz 1990; Scarrow 1991) have addressed the question of the value of the
party on the ground, primarily asking its utility for the party in public office.
The value of the party central office, however, generally has not been
questioned (perhaps because the central office is not generally considered as
a separate face of the party), even though, as we will suggest below, its value
to the other faces is the most problematic.
While not an exhaustive list, we suggest four primary functions for the
party central office. The first is to be the nucleus from which the other two
faces are formed; that is, the party central office may be the core of initial
party activists who go out into the country and organize a party on the
ground that eventually fields candidates who win elections and become the
party in public office. Obviously, this is only one of the possible stories of
party genesis, and it refers to a function whose importance declines as the
party succeeds. A second function of the party central office is to coordinate
national campaigns, which may mean that it supervises or controls the party
on the ground on behalf of the party in public office. Third, and conversely,
on the basis of its permanence, expertise, and location at the seat of government, the party central office may supervise the party in public office on
behalf of the party on the ground. In the same vein, it may aggregate and
articulate the demands of the party on the ground, “producing” the party
congress and acting in place of the congress on a daily basis. Finally, the
party central office may provide a variety of services, such as a party press
or other media of communication, policy research, an efficient fund-raising
organization, and so forth, to the party on the ground and/or in public office.
Evolution of Party Organizations in Europe | 601
Before attempting to apply these ideas, a final qualification is appropriate. Although we are treating the three faces as monoliths for analytic
simplicity, one should remember that they not only may be internally diverse
but also may intersect at multiple points. We have already alluded to
differences between members of the party in public office who are and are
not members of the government; to local activists and more passive members; to members of the party central office based in the party in public
office, the party bureaucracy, and the party on the ground. In a fuller
treatment, one would also need to remember the possibility that particular
members of the party in public office may have strong ties (rooted perhaps
in local control over nomination) to the party on the ground, etc.
Using the Three Faces to Explain Party Organization
How can one employ this characterization of party to understand organizational change and adaptation? In particular, bearing in mind that the
relationships among the three faces of party allow for three dyadic relationships, what sorts of questions do we ask about them? Here, we suggest
three classes of questions as particularly relevant.
The first, of course, is to ask about the resources and constraints of each
face, as we have just done, but bearing in mind that these may vary over
time. In some cases, as for example with the ability of the party in public
office to communicate directly with their electoral supporters without the
intermediation of the party central office or the party on the ground which
was made possible by broadcasting, these changes may be exogenous, with
parties perforce adapting to a new situation. In others, however, as with the
introduction of state subventions (which make the financial resource of the
party on the ground relatively less significant), or the changes in
representation on the national executive (perhaps giving the party in public
office relatively greater weight), they are endogenously introduced to institutionalize or attempt to redress a particular balance within the party.
A second set of questions concerns the independence versus the interdependence of the various faces. This has two aspects. On one hand, two
faces of the party may be in constant contact and exchange relationships
with one another or, alternatively, they may work quite autonomously, each
in its own sphere. For example, there may be a single fund-raising drive with
a uniform national appeal in which both the party on the ground and the
party in public office participate and with a mutually agreed division of the
proceeds or alternatively, each may raise its own funds and control their
disbursement without regard to what the other is doing. On the other hand, to
the extent that there is interdependence between the two faces, the
602 | Richard S. Katz and Peter Mair
relationship may be characterized by mutual influence and accommodation,
or alternatively by the dominance of one face over the other.
Finally, as already suggested in the discussion of party national executives, one may ask the degree to which the faces are distinct versus the
degree to which they overlap.
Before applying this approach to the analysis of recent changes in party
organizations using the data that we have been collecting together with a
team of country collaborators, we can demonstrate its utility by using it to
characterize the three major types of party (cadre party, mass party of integration, and catch-all party) most widely discussed in the literature.
The Cadre Party
The cadre party is characterized by a strong overlap between the party
on the ground and the party in public office. In the pure type, each individual
MP is, from the perspective of the party on the ground of his or her own
constituency, “one of us.” There is not so much a division between leaders
and followers as there is a division of labor within the party on the ground,
which is, by definition, entirely made up of leaders.
The MP essentially combines the roles of member of parliament and
congress delegate, with the parliamentary party in effect serving as the party
congress as well. Resources, however, are monopolized by the party on the
ground, that is by the caucuses of local notables that put up one of their
number as candidate and then support him or her with their private resources. In many cases, the primary source of political capital is the candidate him- or herself. In this situation, party discipline is hard to maintain,
largely because those in control of resources do not want it.
With a cadre party, there is little need, or desire, for a party central
office. Campaigns are local affairs, centering around the mobilization of
local and personal clienteles; thus coordination is not necessary and intervention from outside is more likely to be regarded as interference than as
assistance. While there may be a national executive, and a central headquarters that provides some services to the party on the ground or the party
in public office, they have little independent access to resources, and because they are so dependent on the sufferance of others, they have little
independent weight in the party.
The Mass Party of Integration
In its “genetic myth,” the mass party of integration begins without
either a party on the ground or a party in government. Instead, an initial
group of organizers forms a “central office” which then goes about creating
Evolution of Party Organizations in Europe | 603
the other two faces of the party. While this myth does not completely
describe the genesis of any particular party, just as there are no real parties
that conform totally to the ideal type of the mass party of integration, the
intervention of central leadership is always a necessary catalyst in turning a
mass into a movement or party, and those parties that we usually identify as
mass parties of integration usually went through a process approximating
that in the genetic myth. In particular, even if the party central office was
originated by a few out-cast MPs, they constituted themselves as an extraparliamentary organization, and then went about recruiting members.
The mass party of integration is primarily the child of expansion of
citizenship and participation beyond the limited social basis that formed the
natural home for the cadre party. The mass party arose primarily among the
newly activated, and often unenfranchised, elements of society in their
(ultimately successful) struggle to gain a voice in, and eventually control
over, the ruling structures of the state. It relied on quantity of
members/supporters, attempting to make up in many small membership
subscriptions for what it lacked in individual patronage, to make up in
collective action for what it lacked in individual influence, and to make up
through a party press and other party controlled channels of communication
for what it lacked in access to the commercial press controlled by its
political opponents. Naturally, this required organization.
The strategy of encapsulation is both a response to the need for
mobilization and organization and one of its causes. On one hand, by integrating the citizen into a network of groups which attend to all the needs of
life—news, insurance, union representation, social activities—the party both
provides itself with a mechanism for mobilizing its supporters and a way of
insulating them from alternative influences. On the other hand, the strategy
of encapsulation requires that the party in fact produce that panoply of
organizations and then mobilize its potential supporters to join them.
Underlying the mass party is not only an organizational strategy but
also a distinctive conception of democracy and of the role of a political party
within it. The mass party emphasizes the representation of a particular social
constituency, and its authorization to do so on the basis of the internally
democratic nature of the party itself. Thus, the party congress, as the
representative institution not simply of the party on the ground but (in theory
synonymously) of the politically active portion of the entire social segment
that the party represents, ought to be the supreme decision-making body in
the party and the source, along with electoral success, of democratic
legitimacy for the party as a whole. Elections, moreover, are seen as contests
not between independent candidates competing for the favor of local
constituencies, but between representative teams with alternative programs.
604 | Richard S. Katz and Peter Mair
In particular, the candidates of the mass party are the agents of the party
congress, pledged to support the coherent program that the congress has
enunciated, and competing for a national mandate to put that program into
action. To be able to claim a national mandate, however, requires that the
party mount a nationally coherent campaign, and this as well requires
organization.
In theory, the party in public office is the agent of the party on the
ground, as embodied in the congress. Since the congress cannot be in continuous session, however, it elects a standing committee, or executive, to act in
its place, both in articulating party policy and elaborating on the party’s
electoral mandate and in supervising and directing the party in public office.
Moreover, this executive requires a strong and well endowed central office
in order to perform the various organizational and coordinating functions
that the mass party model implies.
In the original mass party of integration, the balance of resources
clearly favors the party central office, as the coordinator and controller of the
party on the ground. The party in public office is relatively weak, in part
because it is initially very small. At the same time, the party on the ground
and party central office control the resources required by members of the
party in public office if they are to win elections and so stay in office.
The Catch All Party
The third of the major types of party organization, the catch-all
party, arose as a response to the mass party of integration. Essentially that
response had three underlying roots. The first was the success of the mass
party in elections, and especially its success in altering the situation under
which elections were held by expanding the suffrage so that electorates numbered in the millions rather than the thousands. Under these circumstances,
the informal networks of the cadre party were inadequate to canvass, mobilize, and organize supporters. The second was the growing acceptance of the
mass party model of democracy, particularly popular control of government
through choice among unified national parties, even among segments of
society that had traditionally supported the cadre parties. If these two primarily affected the existing or might-have-been cadre parties, the third root
applied (and continues to apply) particularly to the mass parties. This was
that electoral success altered the balance of resources within the mass party
itself, in particular strengthening the hand of the party in public office.
While the success of the mass party led to adaptation by its competitors, they did not adopt the mass party model root and branch. First, they
could not accept the idea that parties exist to represent well defined
Evolution of Party Organizations in Europe | 605
segments of society, because the segments which would have been left to
them (farmers, industrialists, etc.) were obviously and increasingly permanent minorities. Similarly, the idea that the party on the ground, or a party
central office whether independent or primarily responsive to the party on
the ground, ought to be dominant was unappealing to those already established in government. Further, while they needed to organize and mobilize
electoral supporters, they were not so dependent on them for material
resources.
These considerations lead the catch all party to differ organizationally
from the mass party of integration, but organizational differences also stem
from differences in conceptions of the role of parties in democratic systems.
Rather than seeing party as the agent of a particular organized segment of
society, the catch all party model sees party as a broker between autonomous
social groupings and the state. The catch all party defines its constituency
electorally—it is those people who voted for it, or might be enticed to vote
for it in the future—rather than socially and culturally. The party in public
office is seen as an independent entrepreneur responsible to the electorate
rather than to the party on the ground, the party central office, or the party
congress.
The party on the ground remains necessary within the catch all party
model (at least in its initial stages) for several reasons. Just as in the mass
party model, it is a source of important material resources and a channel of
communication between the party in public office and the electorate. The
prevalence of the mass party conception of democracy makes it convenient
for the party in public office to have a mass organization which it can claim
to represent and to which it can claim to be accountable. But the word
“claim” is central to this formulation. That is, the party in public office
wants the appearance, but not the reality, of a strong party on the ground.
This difference is reflected as well in the party central office. So long
as campaigns remain labor intensive and membership contributions remain
an important resource, the party central office remains important, but there is
a strong pressure by the party in public office to domesticate it, particularly
in those parties that evolve toward the catch all model from the mass party
model. (In the case of parties evolving from the cadre model, the central
office is, of course, already subservient.)
Into the Present
Having demonstrated how the three faces framework can be used to
illuminate the development of party organizations in the past, we now turn to
illustrate its application to current developments.
606 | Richard S. Katz and Peter Mair
Indicators of Power
Three dimensions of power are of crucial importance in assessing the
relationships among the three faces of party. The first concerns the numbers
and disposition of the party staff and professional bureaucracy, and the
extent to which these are biased in favor of the party in public office, on the
one hand, or the party central office, on the other. (A heavily decentralized
party bureaucracy may also, of course, reflect a bias towards the party on the
ground.) The second is of specific relevance to the conception of the party
central office as a battleground between the other two faces, and concerns
the extent to which the national organs of the party reflect a bias towards the
representation of the party on the ground or that of the party in public office.
The third dimension of power, which can be related to both of the above,
concerns the intra-party decision-making structure, and the extent to which
any of the three faces may enjoy an authoritative say in matters such as the
formulation of party policy and strategy or the selection of candidates for
public office.
The most obvious of these dimensions is that concerning the numbers
and disposition of the party bureaucracy. Analysis of this dimension faces
two problems, however. In the first place, the resources of the party in public
office may not be visible in pure party terms, especially when the party in
question includes a governing as well as a parliamentary face, and when key
staff are appointed to positions in the public, as opposed to the party,
bureaucracy. In 1993, for example, in the wake of the formation of the new
Fianna Fàil-Labour coalition government in the Irish Republic, it was
estimated that the various ministers had appointed 135 personal staff to their
public offices, at an estimated annual cost of IR£ 3 million (Irish Times 20
February 1993). While this sum dwarfs the expenditure on salaries for the
central office staff for both coalition parties (IR£ 321,000 in 1990—Farrell
1992, 449-50), the bias which it may reflect clearly could not be gauged
solely through an analysis of the parties’ own records.
The second problem concerns the actual responsibilities of the party
bureaucracy. In some cases, for example, parties may derive a substantial
proportion of their income from state subventions to the parliamentary party
(see below), that is, to the party in public office. Indeed, in some countries,
parliamentary subventions remain the only source of state funding. In such
circumstances, the party central office may lack the resources to employ its
own independent staff, and hence in practice those party bureaucrats who are
funded through state subventions in order to facilitate the work of the party
in public office may actually end up working for the central office.
Conversely, bureaucrats who are formally employed by the party central
Evolution of Party Organizations in Europe | 607
office sometimes actually work in the offices of the party in public office.
Knowledge of the formal disposition of the party bureaucracy may not therefore provide an adequate guide to the real bias in organizational resources.
We are on surer ground in relation to the second dimension, which concerns the bias in the pattern of representation in the national executive committee or committees which constitute a major component of the party
central office. As noted above, the members of these committees may be
elected and/or appointed in a variety of ways, whether by the party congress,
the party in public office, or the various affiliated organizations. Some of
these members may be the mandated delegates of any of these constituencies, while others participate on an ex-officio basis. Either way, these
persons are often crucial to the decision-making procedures in the party, in
that they are typically responsible for the day-to-day functioning of the party
between congresses or between elections. The extent to which these committees are dominated by the representatives of the party on the ground, as
against the representatives of the party in public office, can therefore indicate the extent to which either element is seen as the key actor inside the
party as a whole.
Even here, however, any assessment is complicated by a number of key
problems, two of which merit particular attention. In the first place, an
analysis of the bodies formally represented in a committee may have to be
supplemented by an analysis of the individual representatives themselves.
The National Executive Committee (NEC) of the British Labour Party, for
example, is almost entirely constituted by representatives of the local constituencies and of the affiliated trade unions, all of whom are elected by the
annual conference of the party. The party in public office formally is represented only through the inclusion in the NEC of the leader and deputy leader
of the party, both of whom are selected by an electoral college which is itself
only partly constituted by the members of the parliamentary party. Here,
then, we can see a bias towards the party on the ground. In practice,
however, the actual persons who are elected as representatives of the local
parties within the NEC are almost exclusively members, and are often
among the small leadership elite (e.g., ministers or potential ministers), of
the parliamentary party. In practice, therefore, the formal representation of
the party on the ground may in this case be associated with effective dominance by the party in public office.
In the second place, if a party’s national executive committee itself
lacks real power and authority within the party, then an analysis of changes
in its composition may tell us little about the shifting organizational bias
within the party. Thus, for example, while the party membership might
have the exclusive right to select the members of the national executive
608 | Richard S. Katz and Peter Mair
committee, a factor which would indicate a profound organizational bias
towards the party on the ground, the committee itself might in practice enjoy
no real control over the party in public office, thus indicating a bias towards
the latter.
In this sense, any such assessment should properly be complemented by
an analysis of the third dimension indicated above, that is, the actual
decision-making structure. A full analysis should therefore assess the locus
of authoritative decision-making in matters relating to the party policy, party
strategy, party discipline, and party rules. Does the chain of authority run
from the party on the ground to the party central office and then to the party
in public office, or is it precisely the other way around? The problem with
these two scenarios, however, is that each assumes that the party central
office is subordinate. A third possibility is that the party central office
acquires its own independent authority. For example, when the party central
office enjoys the right of final approval of the lists of candidates to be
nominated for election, not only does it curb the independence of the party
on the ground, but it also gains a potential sanction against indiscipline in
the party in public office. Final approval of the party election program offers
it a similar dual advantage.
Party Central Office v. Party in Public Office
A full assessment of the relevance of the three faces of party organization to the understanding of processes of organizational change and adaptation clearly requires the sort of complex analysis which is beyond the
scope of this brief paper. Rather than focusing on all of the possible interrelationships, we will therefore limit our concluding discussion only to the
(potential) conflict between the party central office and the party in public
office. As we have argued elsewhere (Katz and Mair 1991, 1992a), there is
an increasing tendency for the party in public office to be the dominant of
the three faces, a development which may well lead to conflicts with the
party central office. Moreover, from the wider literature we can also assume
at least the possibility that the interests of the party in public office conflict
with those of the party activists and militants, which, in turn, may have a
major influence on the composition of the national executive, and hence on
the party central office. For both these reasons, therefore, we can anticipate
that the party in public office will increasingly attempt to assert its autonomy
of, or even its control over, the party central office.
In pursuing such a goal, the party in public office may follow any of
three distinct strategies, each of which involves one of the dimensions of
power identified above. In the first place, it may push for the introduction
Evolution of Party Organizations in Europe | 609
or expansion of state subventions for the parties in parliament, and so build
up its own independent resources and bureaucracy. Second, it may seek to
exert more authority over the party central office by increasing its own voice
on the national executive committees. Third, it can seek to reduce the weight
accorded to the party central office in the intra-party decision-making
structure, and to assert its own individual autonomy. This it might do
through changes in the party rules, or alternatively, by simply ignoring the
party central office, and appealing to the general population, or to the
(normally largely passive) membership at large.1
Rather than assess all three strategies, however, we will focus instead
only on the first two, and we will concentrate specifically on shifts in the
balance of resources available to the party in central office, on the one hand,
and the party in public office, on the other, as well as on the extent to which
the formal pattern of representation at the national executive level reflects an
increasing voice for the party in public office. The data which we cite are
drawn from the various country studies in Katz and Mair (1992b).
Balance of Resources
We have data regarding two resources that may bear on the potential
conflict between the party central office and the party in public office: state
subventions and staff. Insofar as we anticipate that the party in public office
is seeking to assert its autonomy of, or control over the party central office,
we hypothesize that it will be increasingly favored by the distribution of
these resources.
As far as state subventions are concerned, there is one clear pattern
which emerges and which emphasizes the weight of the party in public
office: in four of the eleven European countries for which data are available,
subventions are provided for the party in public office, with, as yet, no
subventions for the party central office (Ireland, Italy, Netherlands, UK); in
five other countries, subventions were first introduced for the party in public
office, and only later for the party in central office (Austria, Belgium, Denmark, Germany, Norway); in two countries, subventions for both faces of
party were introduced simultaneously (Finland, Sweden); and in no country
for which data are available were subsidies first introduced for the party
central office. Moreover, looking at the most recent data, in only three countries is the subvention to the party central office greater than that to the party
in public office: Austria, where the ratio is in excess of 3:1, Finland, where
the ratio is more than 6:1, and Sweden, where the balance only marginally
favors the party central office.
610 | Richard S. Katz and Peter Mair
In these terms at least, the party in public office has appeared to acquire
substantial additional and independent resources. That said, the fact that in a
number of countries the party central offices began to receive separate
subventions after public office subventions were already in place (Austria,
Belgium, Denmark, Germany and Norway) suggests at least a partial redress
of the balance in their favor, and may in fact indicate a shift in resources
away from the party in public office.
As far as party staff and bureaucracy are concerned, two patterns are
discernible in the nine countries (see Table 1). In the first place, the ratio of
central office staff to public office (parliamentary) staff has declined quite
dramatically over time, a trend which is probably partly the result of the
early introduction of state subventions to the party in public office. In the
first year for which data are available, for example, an average of almost
four times as many staff were employed by the party central offices than
were employed by the various parliamentary parties. In the most recent year,
by contrast, there were fewer than twice as many staff in the central offices.
This shift in the balance of staffing is most dramatic in Denmark and
Ireland, where in both cases some three times as many staff were originally
employed in central office, and where now central office staff are actually in
the minority. In the Netherlands, by contrast, the shift has been only
marginal. In general, however, the direction of movement is clearly and
consistently in favor of the party in public office.
Table 1. Ratio of Central Office Staff to Parliamentary Party Staff*
Austria
Denmark
Finland
Germany
Ireland
Italy
Netherlands
Norway
Sweden
Mean
*
Time 1a
Time 2a
Relative Change (%)
3.9 : 1
2.7 : 1
9.4 : 1
0.3 : 1
3.3 : 1
5.5 : 1
0.7 : 1
3.0 : 1
5.9 : 1
3.9 : 1
2.2 : 1
0.6 : 1
4.1 : 1
0.1 : 1
0.6 : 1
3.1 : 1
0.6 : 1
2.0 : 1
2.3 : 1
1.7 : 1
-43.5
-77.8
-56.4
-66.7
-81.8
-43.6
-14.3
-33.3
-44.1
-56.4
Average per country.
Refers to the earliest (Time 1) and most recent (Time 2) years for which data are
available.
a
Evolution of Party Organizations in Europe | 611
Notwithstanding this, the second pattern is a continuing bias towards
the party central office in absolute terms. Indeed, as noted, the party central
offices now employ an average of almost twice as many staff as the parliamentary parties. This is not true for each individual country, however,
and the national patterns are quite varied. Even in the earlier period, for
example, both Germany and the Netherlands were characterized by a bias
towards the parliamentary party; in Finland, on the other hand, central office
staff outnumbered parliamentary party staff by more than nine to one. By the
later period, Denmark and Ireland had joined Germany and the Netherlands
with a balance favoring the parliamentary party staff, whereas in the
remaining five countries the bias, while reduced, continued to favor the
central office.
Composition of National Executives
It is far from easy to present a general picture of the shifting balances
of representation on the national executive committees of the various European parties. The pattern not only varies substantially across countries and
over time, but also across the different parties within individual countries,
with rule changes and organizational restructuring in the past three decades
marked more by their frequency than by their absence. Accordingly, rather
than presenting a systematic cross-national comparison over time, this discussion will focus simply on some illustrative examples of the sorts of
changes which have taken place since the early 1960s among a set of 35
parties in nine countries.
A large number of parties specify the number of members and/or representatives of the party in public office who are entitled to membership of the
national executives. More significantly, taking cognizance of the possibility
that even those who formally represent the party on the ground may actually
be drawn in practice from the ranks of the party in public office, a handful of
these parties are also very firm in ruling that even in practice this number
should never exceed a stated maximum. Any restriction of the latter type is
an especially important indicator of the desired balance between the
different faces, and is clearly intended to preserve the independent voice of
the party on the ground.
Such restrictions are particularly common in the Netherlands, where
maximum limits on the number of MPs and public office holders have been
established by parties as diverse as the christian-democratic CDA, the liberal
VVD, and the social-democratic PvdA. Since the foundation of the CDA in
1980, for example, the rules regarding the composition of the party’s
national executive (partijbestuur) state that no member is allowed at the
612 | Richard S. Katz and Peter Mair
same time to be a member of parliament, a minister or secretary of state, or a
provincial governor. All elements of the party in public office have therefore
been kept at one remove from the party in central office. The national
executive of the PvdA, on the other hand, is more open to the party in public
office, although here also there are restrictions which have been
strengthened over time. Thus in 1960, it was stipulated that no more than
half of the (19 to 25 member) national executive could be MPs, whereas by
1969, a maximum of seven (of the then 21) members could be MPs. The
PvdA also ruled in 1969 that a maximum of three members of the (9 to 11
member) permanent committee could be MPs, whereas in 1960 no such rule
existed. Nonetheless, the party was clearly intent on having at least some
representation from the members in public office, since it also (and
unusually) included a stipulation regarding the minimum number of MPs
who should be included in the national executive. The VVD, on the other
hand, has moved in the opposite direction to the PvdA, initially stating that a
maximum of three and later four of the (21 to 28 member) national executive
(hoofdbestuur) could be MPs, and later, in the context of a reduced overall
number of members (11 to 13), removing this restriction.
Other examples of parties which are characterized by similar stipulations regarding the maximum number of members from the party in public
office include the Danish Socialist People’s Party (SF), which initially
included no specific rules regarding the inclusion of MPs on either its
national committee (Hovedbestyrelsen) or national executive (Forretningsudvalg), but which later, in 1965, limited their number to a maximum of
five out of 33 in the former body, and to three out of nine in the latter body;
the Finnish National Coalition (KOK), which also initially included no rule
regarding the inclusion of MPs, but which introduced in 1967 the stipulation
that a majority of the members of the national executive (hallitus) may not
be members of the parliamentary party; and the Irish Fianna Fàil (FF),
which, from the beginning, stipulated that there could only be a maximum of
five MPs or Senators in its traditional “Committee of 15,” and that none of
the 50 or so constituency representatives on its broader national executive
committee (Àrd comhairle) could be MPs. In this case, however, the
restriction was balanced by subsequent rules (in 1971) which broadened the
executive committee’s membership to include three co-opted members of
the government or front bench, as well as five backbench members of the
parliamentary party.
These six parties are particularly important insofar as they place limits
not only on the formal representation of the party in public office, which is
in fact quite common among all of the parties which we have examined, but
also because they are illustrative of a relatively significant group of parties
Evolution of Party Organizations in Europe | 613
which, in one way or another, are explicitly concerned to limit this representation in practice. Moreover, it also should be emphasized that there is
little sign of a softening of this emphasis on the need to give practical voice
to the party on the ground. Indeed, of the six examples cited here, only two,
the VVD and FF, can be regarded as having eased the limits on the party in
public office, while three others, the PvdA, SF, and KOK, have actually
adopted a more restrictive attitude.
These latter three parties seem exceptional, however, in that the more
general trend does tend to reflect a gradual strengthening of the position of
public office holders. Thus, in a wide range of parties, including the FPÖ in
Austria, the PRL/PVV and Volksunie in Belgium, the RV in Denmark, the
FDP in Germany, the Labour Party in Ireland, and the Conservatives in
Norway, the balance of representation on the various national executives is
now more likely than before to favor the representation of the party in public
office.
In the FPÖ, for example, the rules concerning the representation of
the party in public office on the 40-member national executive (Bundesparteileitung) initially stated only that it should include the deputy chairman of the parliamentary fraktion; by 1972, however, the now 86-member
body was defined as including all the party MPs in both the Bundesrat and
Nationalrat. In the case of the Norwegian Høyre, the rules concerning the
composition of one of the two national executive bodies, the arbeidutvalegt,
initially included no stipulation regarding the party in public office, but
were modified (in 1962 and 1970) to ensure that both the chairman of the
parliamentary group and the chairman of the government group would be included as members; the rules concerning the party’s second executive body,
the sentralstyret, were also modified, increasing the public office
representation from simply 3 MPs in 1960, to the chairman of the MP group
and five other MPs in 1962, and then to the full executive committee of the
parliamentary party and the party cabinet members in 1970.
The other parties listed above evidence similar trends. In the case of the
Belgian Liberals, for example, the executive committee of the formerly
united PRL/PVV already accorded a substantial weight to the party in public
office, including the fraktion leaders in both houses of parliament, the members of the government, and the ministers of state. By the end of the 1980s,
the two executives of the now linguistically divided party also included all
MPs. In the case of the Volksunie, the rules regarding the composition of the
national executive initially included nothing specific about MPs; by 1989,
they specified the inclusion of four MPs, together with all members of the
government (albeit without voting rights). The pattern in the Danish Social
Liberal Party was similar, the initial absence of any stipulation being
614 | Richard S. Katz and Peter Mair
succeeded by the inclusion of two MPs. In the case of the national council of
the German FDP, the inclusion of all fraktion members in an advisory
capacity was succeeded by rules which reduced this to eleven fraktion members in an advisory capacity but which also added the 7-member fraktion
executive with full voting rights.
Finally, in the case of the Irish Labour Party, an initial representation of
the party leader and deputy leader, together with two MPs on a 25-member
body (the administrative council), was modified to include an additional four
MPs and the chairman of the parliamentary party on a newly-structured 52member body (the general council). In addition, this new body was also
obliged to elect a core 13-member executive committee, which had to
include the party leader, deputy leader, and the chairman of the parliamentary party. That said, it should also be emphasized that from 1989
onwards a new rule stated that the positions of party leader and deputy
leader were to be filled by a postal ballot among all party members, whereas
prior to this it was only MPs who enjoyed the right to vote in leadership
elections (MPs do, however, continue to have the exclusive right to nominate candidates, and both positions can be filled only by MPs).
In general, then, these examples suggest a relatively widespread tendency to strengthen the position of the party in public office on the various
national executive bodies. And except in the few cases where more severe
limits have been placed on what was already a specified maximum level of
representation by public office holders, the trend rarely seems to go against
this face of the party. One important exception, however, is the Austrian
Peoples’ Party (ÖVP), the executive committee (Bundesparteileitung) of
which in 1960 included six members of the parliamentary party, as well as
all government ministers among its 47 members. By 1990, however, following a restructuring, this committee included among its 52 members only the
Federal Chancellor, the President of the Nationalrat, and the Chairman and
Deputy Chairman of the Bundesrat. The parliamentary representation had
therefore been reduced, while the representation of members of the government had effectively ceased.
The Decline of Central Office?
Each of these elements—resources, staffing, and pattern of representation—tends to indicate that greater weight is being given to the party in
public office. We now return to the question raised at the end of Section II,
to speculate about the future of the party in central office as an independent
face of European party organization.
Evolution of Party Organizations in Europe | 615
As the proportion of electorates that are included among party members
declines—as it is doing virtually everywhere in Europe (Katz, Mair et al.
1992)—and as the atypicality of the party activists who dominate local party
organizations and national congresses seems more apparent, the claim of the
central office to legitimacy as the democratically elected representative of an
encapsulated political community might be expected to weaken. If there is
also a tendency among voters to revalue the “public interest” over the
partisan interests of their own social and cultural segments, the legitimacy of
central office should be further reduced.
Party central offices may also be losing the power that flows from the
unique ability to perform crucial functions for the other faces of the party.
Mass media and computer-generated direct mail allow the party in public
office to communicate directly with the electorate and with the party on the
ground without the intervention of an elaborated party organization. As
shown above, the party in public office increasingly has its own staff and its
own financial resources. The members of the party on the ground also have
many alternatives: media to keep informed; interest groups and new social
movements to become involved and to exert pressure. Further, with the
possibility of postal ballots and other forms of direct decision by members,
there is less need for representative institutions, such as the central office
might claim to be, within the party.
As these trends develop, the central office may, indeed, lose its
centrality. Looking at party central office budgets, it appears that there is
still plenty that the central office does. The point is, however, that while the
central office may still be useful, it is not indispensable, because most of the
services it provides can now be secured through alternative means. For the
party in public office, communications services can be bought on the open
market, perhaps at a higher price in money, but without the added costs of
subservience to a party organization whose goal priorities may be quite different from their own. For the members of the party on the ground, as well,
the alternative ways to be involved in national politics may be less confining
and more satisfying, and this may help explain the relative decline in party
membership itself.
The scenario this leads us to hypothesize for Europe is in many ways
one that is already familiar to Americans, although in other respects it
remains quite distinctive. Here it is worth recalling that it is only in the
mass party model that the party central office plays an especially important
role. A decline in importance of central office is not so much a crisis of
party, or a decline of party, as it is a redefinition and reorientation of
party—a crisis of the mass party model, indeed, but not necessarily of party
itself, unless the very definition of party is tied to the mass party model. On
616 | Richard S. Katz and Peter Mair
one hand, the party central office simply becomes a bureaucracy serving the
party in public office, perhaps supplemented or even largely supplanted by
consultants hired on a contract basis, but with no independent authority
except that based on the ability to get politicians to defer to their professional expertise. Moreover, as they channel their political participation
through organizations other than party, the members of the party on the
ground have less interest in a party central office to supervise the party in
public office. In these respects, European parties might come to look like
their American counterparts. At the same time, the exigencies of parliamentary government suggest that the party in public office will remain far
more centralized and disciplined than in the United States.
On the other hand, the party on the ground may become far more
autonomous with regard to local politics. In some countries (e.g., Norway),
this face of the party now receives direct public subventions, and so is not
financially dependent on the central organization. As the party on the ground
becomes less necessary to the national party in public office, the party in
public office has less need of a central office to supervise the local party on
the ground. And although the central office may be able to provide services
to the party on the ground “wholesale,” the members, like the party in public
office, have the possibility of alternative suppliers. Organizationally, the
projection would be for stratarchy to replace hierarchy. More generally, it is
a projection of the party central office transformed from a potential power
center into a service organization.
NOTE
1
Discussions of party organization which are based on a simple dichotomous division between
leaders and followers, or between parliamentary and extra-parliamentary party, necessarily assume
that any such circumventionist strategy involves an appeal to the population at large, or to the voters
at large. However, once we allow for the existence of three separate faces, and recognize that the
party on the ground consists both of activists (who are likely to occupy local party offices, serve as
congress delegates, etc.) and of more passive members, we can also envisage circumventionist
strategies which are aimed at the membership at large. Indeed, it is precisely this strategy which
seems to be increasingly characteristic of a large number of contemporary European parties.
REFERENCES
Aberbach, Joel et al. 1981. Bureaucrats and Politicians in Western Democracies. Cambridge:
Harvard University Press.
Andeweg, R.B. 1992. Executive-Legislative Relations in the Netherlands: Consecutive and CoExisting Patterns. Legislative Studies Quarterly 27: 161-182.
Evolution of Party Organizations in Europe | 617
Daalder, Hans. 1983. The Comparative Study of Parties and Party Systems: An Overview. In Hans
Daalder and Peter Mair, eds., Western European Party Systems. London: Sage.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper.
Duverger, Maurice. 1951. Les Partis Politiques. Paris: Colin.
Farrell, David. 1992. Ireland. In Richard S. Katz and Peter Mair, eds., Party Organizations: A Data
Handbook. London: Sage.
Katz, Richard S. 1987. Party Government and Its Alternatives. In Richard S. Katz, ed., Party
Governments: European and American Experiences. Berlin: de Gruyter.
__________. 1990. Party as Linkage: A Vestigial Function? European Journal of Political Research
18: 143-161.
__________ and Peter Mair. 1991. Changing Models of Party Organization: The Emergence of the
Cartel Party. Paper prepared for the European Consortium for Political Research Joint
Sessions of Workshops, Limerick, Ireland.
__________. 1992a. Varieties of Convergence and Patterns of Incorporation in West European Party
Systems. Paper prepared for the European Consortium for Political Research Joint Sessions of
Workshops, Leiden, The Netherlands.
__________, eds. 1992b. Party Organizations: A Data Handbook. London: Sage.
__________. Forthcoming. The Organisation of Parties in Western Democracies: National Patterns.
London: Sage.
__________, et al. 1992. The Membership of Political Parties in European Democracies, 1960-1990.
European Journal of Political Research 22: 329-345.
Kirchheimer, Otto. 1966. The Transformation of Western European Party Systems. In Joseph
LaPalombara and Myron Weiner, eds., Political Parties and Political Development.
Princeton: Princeton University Press.
Laver, Michael and Norman Schofield. 1990. Multiparty Governments: The Politics of Coalition in
Europe. Oxford: Oxford University Press.
Lewis-Beck, Michael. 1988. Economics & Elections. Ann Arbor: University of Michigan Press.
Panebianco, Angelo. 1988. Political Parties: Organization and Power. Cambridge: Cambridge
University Press.
Scarrow, Susan E. 1991. Organizing for Victory: Political Party Members and Party Organizing
Strategies in Great Britain and West Germany, 1945-1989. Ph.D. Dissertation, Yale University, 1991.
618 | Richard S. Katz and Peter Mair